Apple is researching heart rate measurement using artificial intelligence (AI), potentially transforming AirPods into sophisticated health monitoring devices. 9to5Mac recently reported that Apple’s research into AI-driven heart rate estimation has yielded promising results, with accuracy levels comparable to those of current methods.
Apple’s research team employed six AI models, trained on audio and speech data, to analyze heart sound recordings. Apple’s proprietary CLAP (Contrastive Language-Audio Pretraining) model showed particularly impressive performance, bolstering the case for AI-based cardiac monitoring. The study used a dataset of 23,000 heart sound samples to train the AI in estimating heart rates.
Interestingly, the research revealed that bigger isn’t always better when it comes to AI models for this application. Shallow or mid-level layers proved more effective at heart rate estimation, a finding that could significantly influence Apple’s development of AI-powered health features.
The study suggests that marrying cutting-edge AI with traditional signal processing techniques could boost the reliability of heart rate measurements. Looking ahead, the team aims to develop more lightweight AI models suitable for low-power devices. While it’s unclear whether this research will directly lead to new consumer products, it certainly points to a future where your AirPods might keep tabs on your ticker.